Automated Peering

We use PeeringDB and CAIDA datasets to identify possible peering points for requester and candidate ISPs. We estimate candidate ISP's traffic matrix and consider ISPs' internal policies to generate a list of acceptable peering contracts.

We have developed a simple web application, which can be accessed at https://metapeering.net/, that allows ASes to select their desired Points of Presence (PoP) locations and analyze the possibility of peering with any other AS.

This repository contains the code files for peer selection algorithm, as well as the web application.

Note: DO NOT make this repository public before removing the web app code.

Running and developing the app locally:

MAC Tutorial:

Python version required: 3.x

1. Clone this repository, and open a terminal in the same folder.
2. To create a new virtual python environment python3 venv metapeering
3. To activate the created virtual environment, run source metapeering/bin/activate
4. To install all the dependencies, run pip install -r requirements.txt

Windows Tutorial:

Python version required: 3.x

1. Clone this repository, and open a terminal in the same folder.
2. To create a new virtual python environment python -m venv metapeering
3. Now cd into the virtual environment scripts folder cd metapeering/scripts
4. Run the script to activate using activate
5. Cd back cd ../..
6. If you have run locally before: Uninstall your site-packages using pip uninstall -r requirements.txt
7. Install packages using pip install -r requirements.txt
8. Now start the local server using flask run

If running on Windows, make sure to change the syntax of some commands in app/routes/home.py and app/routes/custom.py, specifically the mkdir commands which work in MAC but not in Windows, they need to have backslashes. Revert these changes before pushing to the webserver.

Also, by default Flask does not have live updates, after making a change you will have to use Ctrl+C in Windows or Command+C in MAC to terminate the local server, and use flask run again to see the updated website.

AWS Elastic Beanstalk Deployment Steps

1. Create an Elastic Beanstalk Environment
2. Create CI/CD Pipeline
3. Test SSL Certificate Generation
4. Configure Route 53
5. Update Elastic Beanstalk Environment Configuration

Create an Elastic Beanstalk Environment

Variable name	Leave empty
AWS_S3_URL	no
AWS_STORAGE_BUCKET_NAME	no
DATABASE_URI	no
DOMAIN_NAMES	yes
FORCE_SSL	yes

Items that can be empty/null should be defined but without an entered value.

Consult internal documentation for values of environment variables that should not be left empty.

By default, PYTHONPATH may be among the environment variables. Do not modify this value.

Create CI/CD Pipeline

Create a CI/CD pipeline in AWS CodePipeline. The source should be the GitHub repository and a chosen branch. The Deploy stage should set the Elastic Beanstalk environment as the target.

Test SSL Certificate Generation

Before connecting a domain with Amazon Route 53, you can use the default Elastic Beanstalk CNAME that is created with an Elastic Beanstalk environment to test SSL certificate generation. Ex: metapeering-test-env.eba-<env_id>.<region>.elasticbeanstalk.com. Place the CNAME for your environment in DOMAIN_NAMES and set FORCE_SSL to true. This should trigger an automated update of the environment that includes the generation of an SSL certificate for the given domain.

Configure Route 53

Configure Route 53 with a hosted zone for your domain. Create a CNAME record for your domain that targets the CNAME provided by Elastic Beanstalk for your environment.

Update Elastic Beanstalk Configuration

Update DOMAIN_NAMES to include your domain from Route 53. Domains should be separated by a single comma (no space) in the value for DOMAIN_NAMES.